RCD Test on lighting circuit protected by RCBOs

[traineespark]

Hi everyone, still training so please bear with me.

What is the correct way to carry out an RCD test on a lighting circuit protected by a BS61009 Type B RCBO?

I've looked in the regs, OSG and all of my course material but still no actual definitive answer so thought I'd ask here.

The reason I ask is that everything I've been taught to date as said that we only test live where it is save to do so, which is why Zs on lighting circuits is normally calculated rather than measured. That's all fine but you can't calculate the RCD test results.

Thanks in advance

 

[Silly Sausage]

The definitive answer to testing RCDs, is that you carry the tests out at the device with the outgoing circuit cables disconnected. (there are possibly some varying manufacturers variations)

So, you'll have to power down first to make your disconnections.

 

[westward10]

It is not possible to test the rcbo with no mains power, it is one of the live tests.

 

[traineespark]

Thanks for your replies. So you are saying that to carry out the RCD test I should be doing the following (just checking I've understood correctly):-

1. Switch off main switch.
2. Disconnect L1 and Ln on outgoing side of RCBO
3. Power up
4. At the RCBO test between the L & N terminals and earth bar

Thanks again

 

[westward10]

Yes.

 

[GMES]

This trips a lot of people up still to this day but as others have said test at the load side of rcbo or rcd. You are testing the device not the lights.
Edit
And yes make sure loads are disconnected.

 

[traineespark]

Great stuff, thanks gents. You'll not believe how much trouble I've had getting a straight answer to this!

 

[Andy78]

The test procedure is described in the RCD section of the OSG. There is also a very useful general RCD resource here.....
RCD Handbook (Dec 2010) - http://www.beama.org.uk/resourceLibrary/rcd-handbook---dec-2010-.html
Well worth downloading for future reference

 

[Leesparkykent]

The test procedure is described in the RCD section of the OSG. There is also a very useful general RCD resource here.....
RCD Handbook (Dec 2010) - http://www.beama.org.uk/resourceLibrary/rcd-handbook---dec-2010-.html
Well worth downloading for future reference

He's back. Where you been hiding mate? lol

 

[Andy78]

Hello I've been away for notably long ? Just not logged on for a few weeks due to being busy really.

 

[Deleted member 26818]

One of the problems I have with conducting RCD tests at the board, is that they do not prove tha RCD will provide the required protection.
To my mind, an RCD installed at a board to provide protection for socket-outlets should be tested at the socket-outlets.
An RCD installed at the board to provide protection for cables concealed in walls should be tested beyond wherever the cables are concealed.
An RCD installed to provide protection for circuits of a location containing a bath or shower, should be tested in the location.

 

[westward10]

It should make little difference.

 

[Midwest]

Look, about a year ago we had the same debate. At great expense to myself (if I say so myself!), I emailed several manufactures and organisations and the overwhelming majority were in favour of carrying out an RCD test at the end of the circuit.

I would have posted a link, but not intelligent enough to use the search facility here But you argue all you want, I will continue to test at the circuit end, or as near as.

For the OP, its one of the live tests. Testing a socket is easy enough, testing an RCBO inside a CU or at a lighting point, both have an element of risk. But you should be capable of doing given your training.

 

[Octopus]

Testing at a light - easier if you invest in kewtech adapters ......

 

[hightower]

Testing at a light - easier if you invest in kewtech adapters ......

Still need to unscrew a plastic pendant to access the earth block (if you're a peasant like me with a 3-lead tester!)

 

[Deleted member 26818]

Plug line and neutral probes into the adaptor, unscrew the rose cover, clip the earth lead to the earth block with your croc clip.

 

[telectrix]

i'm with spin on this. RCD might perform perfectly when loads are dissed, but a neutral fault somewhere on the installation could prevent it from operating. the whole purpose of a RCD is to protect against accidental touching of a live part anywhere on the installation. therefore it should be tested as a real life situation, not dissed from the circuits. that's like something working in the lab. but failing in service. tin hat, where art thou?

 

[freddo]

I agree with the above, but when say replacing a DB I'll often fill it with the required RCBOs power up and test them all before connecting any of the outgoing cables. Hopefully insulation tests will identify any neutral problems, but some electronic equipment can upset RCDs, how far do you go, plug in and switch on every load in the installation before testing?

 

[Octopus]

Still need to unscrew a plastic pendant to access the earth block (if you're a peasant like me with a 3-lead tester!)

Cripes.... if you can't do this safely, you shouldn't be meddling with lecky!

 

[hightower]

Cripes.... if you can't do this safely, you shouldn't be meddling with lecky!

Just noting that for all the effort of getting the screw-ins it still means you have to access near live terminals. That said, it's a nightmare holding three leads in the air while pressing a button so those screw-ins are worth their weight.

 

[freddo]

The cap pins fell off of my BC one, not happy considering the cost of the things!

 

[Octopus]

The cap pins fell off of my BC one, not happy considering the cost of the things!

had mine for about 6 years - still going strong!

 

[Midwest]

I just follow the series of tests as per BS7671. Can't understand, having completed the dead tests, and securing the cover to the CU, to carry out Zs tests, to then remove the CU lid disconnect the circuits and carry out the RCD test
PS All appliances/lamps etc, disconnected for all of the tests.

 

[Deleted member 9648]

i'm with spin on this. RCD might perform perfectly when loads are dissed, but a neutral fault somewhere on the installation could prevent it from operating. the whole purpose of a RCD is to protect against accidental touching of a live part anywhere on the installation. therefore it should be tested as a real life situation, not dissed from the circuits. that's like something working in the lab. but failing in service. tin hat, where art thou?

The circuit should be IR tested before being energised....if you haven't found a N-E fault before live testing shame on you! ()

I normally test at a convenient point on the circuit rather than at the RCD terminals....simply because it's usually easier depending on connected loads. Newbies need to remember that anything with a capacitor will seriously affect readings and may make a perfectly sound RCD fail the tests.
When it comes to RCBO's it's also the case that some will not respond to a test direct to terminals for some obscure technical reason nobody understands.....these must be tested via the connected circuit,or a couple of short tails need to be connected.

 

[westward10]

The only ones I know of are the user friendly Memshield 2.

 

[Octopus]

Just noting that for all the effort of getting the screw-ins it still means you have to access near live terminals. That said, it's a nightmare holding three leads in the air while pressing a button so those screw-ins are worth their weight.

Why, the ES and SES adaptors are properly insulated and you can plug your insulated leads into them..... or am I missing something?

 

[hightower]

Yeah fair enough

 

[sparksburnout]

RCD might perform perfectly when loads are dissed, but a neutral fault somewhere on the installation could prevent it from operating

Explain how?

 

[Midwest]

Explain how?

Agree with Tel, long story which I previously recounted (RB explained, I think) earth neutral fault, stopped RCD functioning.

 

[sparksburnout]

Agree with Tel, long story which I previously recounted (RB explained, I think) earth neutral fault, stopped RCD functioning.

Can't see how. Also struggling to see the point in testing multiple circuits at the end of each when fed from a common RCD? I was always taught to test an RCD at the board, and have always done so.

 

[westward10]

This came up on another thread about a month or so ago. A neutral earth fault may not trip an rcd if the fault decides to use the neutral and not the cpc as the return path due to the cpc resistance.

 

[Deleted member 26818]

One of the things that will prevent an RCD operating is d.c. current.
This is how the 'no trip' function on MFTs works.
The MFT saturates the coil with d.c. current while it conducts the Zs test.

 

[HandySparks]

When it comes to RCBO's it's also the case that some will not respond to a test direct to terminals for some obscure technical reason nobody understands.....these must be tested via the connected circuit,or a couple of short tails need to be connected.

It's because a few RCBOs have the current sense coil between the cable entry ports and the terminals (above the location of the terminal screws in conventional vertical orientation) and are sensing the current in the first few millimetres of wire outside the terminals. So if you put your test probes on the terminal screws, you're bypassing the coil. If you probe via the cable entry ports, it'll test OK, but this isn't usually possible due to the space available.

 

[Deleted member 9648]

It's because a few RCBOs have the current sense coil between the cable entry ports and the terminals (above the location of the terminal screws in conventional vertical orientation) and are sensing the current in the first few millimetres of wire outside the terminals. So if you put your test probes on the terminal screws, you're bypassing the coil. If you probe via the cable entry ports, it'll test OK, but this isn't usually possible due to the space available.

Ooohh, I thought it was something much more complicated!! Makes sense though,thanks for that.

 

[westward10]

One of the things that will prevent an RCD operating is d.c. current.
This is how the 'no trip' function on MFTs works.
The MFT saturates the coil with d.c. current while it conducts the Zs test.

I think you will find nearly all rcds now sense AC and DC which is why the old style of loop tester like Robin (still using one) which used their DLoc technology not to trip the rcd rarely operate through rcds now, this technology used DC which most rcds basically didn't recognise. Modern loop testers use low current of around 10ma so as not to upset the rcd, a method which often supply gives inaccurate results.

 

[Midwest]

Can't see how. Also struggling to see the point in testing multiple circuits at the end of each when fed from a common RCD? I was always taught to test an RCD at the board, and have always done so.

I don't know how it does, but it did in my case......keeping it short, connecting up CU & found neutral earth fault on a radial. Decided to investigate that later, and put offending cable to one side, except I didn't and mixed up with another circuit. Carried testing and connecting. Got to testing one of the RCD's & it would not trip under test nor via test button. After a lot of head scratching, went back to the beginning of testing again, but after a short while tested the single 'faulty' cable not connected & realised my mistake. Removed the actual faulty cable from CU, everything hunky dory. Not sure of the science, but that's the reality.

Think either Lucien or Richard came up with an explanation.

 

[Deleted member 26818]

I think you will find nearly all rcds now sense AC and DC which is why the old style of loop tester like Robin (still using one) which used their DLoc technology not to trip the rcd rarely operate through rcds now, this technology used DC which most rcds basically didn't recognise. Modern loop testers use low current of around 10ma so as not to upset the rcd, a method which often supply gives inaccurate results.

No, only 'B' type sense d.c. current, which is why they are used where there are solar panels.